19 February
2007

FilmLoop and Alexa - When Fake Rankings Kill Companies

Or why VCs should stick to term sheets and stay out of the datacenter

There's much sturm und drang about the nasty setup and selloff of a little company called FilmLoop in the business community this week. While there can be much debated about liquidation preferences, drag-along rights and sharp practices, there is one issue relevant to the datacenter operator - the claim by some VCs that Alexa rankings are a good validator of an Internet company's audience and future traffic. And now, since we've stepped out of money-land and into datacenter-land, let's examine this assumption a bit more carefully. Is Alexa a good validator of a business, or not?



First, how does Alexa obtain traffic information for its ranking? Since the only way you can honestly obtain information is by monitoring the switches in the datacenter itself and analyzing the traffic, one might assume that Alexa has some deal with datacenter operators who provide them this information. If you thought so, you'd be very wrong. All Alexa does is provide a toolbar that supposedly tracks your favorite sites and reports this information back to them. While this may seem a good substitute, it really is an absurdity, since most people do not use such a toolbar and the small audience who would choose to install such an item is unlikely to be representative of the audience you are claiming to model. It also doesn't work on all browsers (I'm using Firefox right now) or systems like Unix or Macs. So their toolbar user base is definitely not the group who I talk to, even though I have a very good following. It's just silly to even consider it is anything but a joke.


So if it is such a joke, why are some folks so wedded to the notion that it actually means something? Well, while I told you it can't track real traffic, it *can* be used to *fake* traffic on websites that no one really visits. How? Let's say you have a little group of friends and a website that no one cares about. By making sure each of your friends and their friends all install the Alexa toolbar and then visit this site, you can artificially inflate the traffic numbers even though no one else ever visits, because the numbers are extrapolated to a larger audience than exists!


So the sad little story is Alexa is essentially meaningless. It does not track pageviews -- it guesses activity based on a toolbar reporting mechanism instead of visits and packet traffic. It doesn't have enough information to determine actual activity. Good ratings can be faked and real ratings cannot be tracked. As people like Om Malik have observed, anyone who depends on this for a business investment is a complete and utter fool, but these fools exist and entrepreneurs should be forewarned. If poor little FilmLoop was sacrificed on the alter of bogus Alexa traffic rankings after years of hard work and struggle, who will be next?


Posted by lynne : "FilmLoop and Alexa - When Fake Rankings Kill Companies" at 17:02 | link to entry | Comments (1)
<< Strange Friday - Anna Nicole, Jim Gray, Nowak | Main | Moms in Tech, Really? Impossible! >>
Comments
Re: FilmLoop and Alexa - When Fake Rankings Kill Companies

I agree 100%. Apparently there is lots of Alexa-love out there, despite the points you've raised. I guess I just don't understand the people who seem to have faith in all these online audience measurement systems as representative of the actual user base. Maybe they just don't realize the true extent of how easily traffic can be faked, or just don't care.

Posted by: Greg Skinner at March 15,2007 17:29